Sushi’s Outlook on Push Traffic in 2021
The following is a repost from our founder’s blog
Seeing as I mainly run push traffic as well as build out push subscriber lists, I thought people may be interested in my views (conceited I know) on this traffic source for 2021 and so I thought I’d share my thoughts below (also writing things out help me think and crystallize my plans):
Around to Stay and More Like Email
Despite the naysayers I expect push to be around and stay. This is because I think push can be very useful. For instance it serves as a great outreach for businesses that want to reach their customers/prospects regarding their latest sale but not have to deal with the hassle of building out a full fledged app.
And so instead of drawing the analogy to push vs pop I would say it is more alike to email. And just like email used to be, push is also flooded with spam and abuse.
Clamping Down On Spam + Harder to Get Subscribers
Biggest difference between email back in the day and push nowadays is that push is pretty much controlled by one company, Google. Hence they have much more control in controlling things like spam/abuse and you are already seeing this in 2020 with the changes like a quiet UI.
I expect this theme to continue in 2021 where they will introduce client side changes (i.e. in Google Chrome) that make it harder for you to get push subscribers as well as harder for you to reach your push subscribers.
Eventually they will also clamp down on the sending side. Just like how email servers have a reputation score I expect this to eventually get introduced to push sending servers as well.
Improvements To Push Management Softwares (Hopefully)
2020 saw a host of new push software come out and expect that we’ve probably hit the point where we need better software and not more. The biggest issue I expect (or hope) will get addressed is the ability to build funnels for push notifications just like you would for email.
Not sure if anyone has tried but building out a sequential funnel for push on a software like Truepush (this is the one I use most) is very very painful. As far as I know, OneSignal should be fairly similar too.
Hence I expect that we’ll get software that aligns closer to how we manage emails.
In-Page Push Carve Out
I think people will realize in-page push is not push (most do already by now). It is much closer to display ads or interstitial than classic push. Just think of how you see in-page push, so far I’ve been seeing them a lot on streaming sites (which historically is filled with pops/display ads) whereas classic push shows up in the notification section of phones and computers.
The quality of in-page push is usually much worse, hence traffic sources really should be carving out in-page push (I’d rather see them grouped with display or interstitial).
Uncertain: Will It Come To Apple/Safari?
Given Apple’s focus on privacy and what not, I can’t imagine them allowing push unless they decide to build their own version of it (since they pretty much have their own version of everything). However whether this is even worth their time really depends upon how Google handles the governance of push traffic as a whole. So yea I think this one is a bit of a stretch and we will probably need to just stick to in-page push to reach Apple users.
In Conclusion
If it isn’t clear already my view of push is that it will continue its evolution to become more like email.
Spamming people will still work from time to time but not ideal long term. More and more people will begin managing their own push list and there will be more emphasis on creating value with your content as well as sanitising your list properly (just like how you need to manage email bounce and complaint rates).
Similarly my focus in 2021 is still primarily in push but with a longer term view in mind. There will be more effort in managing our own push lists to cater to our audience so as to build “moats” around my business. Hopefully I can find a way to share more of this information with everyone, though honestly just tracking the LTV of a push subscriber makes me do this:
